|
|
|
|
|
by opportune
2655 days ago
|
|
IME there are only two kinds of Ops structures. Either you have an organized, byzantine, inflexible system with way too many developers actually working on ops full time, or you have this kind of system which is more flexible in the short term but becomes an operational nightmare as technical debt accrues and accrues. I'm not sure if there is a way to get something in the middle. The second system is a result of Agile methodology where some dev makes the mistake of telling the PM/management "well, we could get it deployed today if we just do [terribly hacky unmaintainable thing]" and gets the ok because the project is two weeks behind and half the team is going on vacation next month. The first system is the result of the "everything must fit the spec" methodology where you have excessivley long and inflexible iteration cycles, usually because you are an older more enterprisey organization. |
|